Using TrackLab to understand farm animal behavior

TrackLab is the tool for recognition and analysis of spatial behavior in farm animals. After you decide on the most suitable tracking technology, TrackLab software lets you import the collected tracking data in real-time or offline. The data you collect can be visualized, processed and analyzed. Furthermore, you can create interactive systems that respond in real-time to the location or spatial behavior of animals being tracked. The system will automatically quantify parameters like how much time the pigs spent by the water supply, the average speed of the pigs whilst they were moving, how much time they spent lying down, how often they moved, etc.

tracker
TrackLab works with a variety of tracking technologies.
trackers farm animals

Some of the tracking devices that can be worn by farm animals. 

Tracking technology

TrackLab is designed as an open and flexible system, so that the software works with a wide variety of tracking technologies. For outdoor tracking you can use GPS data, either real-time or offline and you can improve the data quality with augmentation services such as EGNOS or WAAS. Indoor tracking can be used in barns or stables.

TrackLab supports a wide variety of indoor tracking solutions, including Ubisense™ ultra-wideband sensors and tags, and EagleEye™ stereo camera sensors. We are also close to releasing new software for automatically tracking the locations of multiple animals using simple video cameras. TrackLab suports both offline import of track files and real-time live data import (in both cases it is able to handle a wide variety of formats).

Data Preparation

As with all measurement systems, location measured using GPS devices has errors associated with it. The size of the error varies according to atmospheric conditions, the location of the subjects (for example, tall buildings can block satellite reception), and movement of the subject. TrackLab incorporates data processing algorithms to remove outliers and smooth the data which specially designed to improve the quality of GPS track data. Furthermore, there is also a powerful function to select and edit individual tracks to improve your data quality.

Visualization

You can explore and view your TrackLab data in a variety of ways: Locations show data for each track point (or GPS fix) including position, speed and acceleration. You can also copy this data to other programs for further analysis and sort and edit the data in this view to remove outliers. Maps show your tracks either in OpenStreetMap or on an image of a floor plan of your barn, which you can import (depending on your type of data). You can pan and zoom the map, and draw points and regions of interest. You can plot the tracks as trajectories (in a variety of styles), or show the sample density as a customizable heat map. The tracks can be played back in a variety of ways, so that both short tracks and those acquired over a long time period can be played back conveniently. Graphs show the value of a measured variable (currently only speed) over time, synchronized with the other views. The data of all the views are synchronized so that you can explore the interaction between, for instance, velocity and location.

The heat map is generated according to the density of the GPS samples. It can be seen that the cows spent a relatively large amount of time in the stall (on the right), and near the silage and water (in the centre) and that when in the field they spent less time at the far end of the field than nearer to the farm buildings.

The colour of the line indicates the speed of the cow at that moment. A. In a pasture.  You can see that the cow is moving slower in the region next to the farm buildings where the water and silage were available.  B. At a wooded site.  Searching behaviour (long flights, high speed) can be distinguished from foraging behaviour (short flights, low speed).

Statistics

You can use the location of your tracked subjects in relation to zones (feeding troughs, inside/outside, etc) and also whether or not they are moving to generate events. You can create an analysis report showing detailed information about each of these events. You can also make an analysis report showing a wide range of variables for each of your tracks, with statistics such as duration, average speed, maximum acceleration, movement statistics (according to thresholds you can set yourself), a variety of parameters describing the path shape, and statistics quantifying the zone visits.

Real-time integration with other software

TrackLab is able to provide its data real-time to other software. For instance, when a cow enters a zone next to a feeding trough, or starts moving, that generates an event, which can be communicated to another program, for instance to open a door to another part of the stall. TrackLab’s real-time integration opens up a whole set of possibilities, and Noldus will be happy to help you configure it to your particular needs.

Specifications & References